Cascading failures in urban infrastructure systems: A comprehensive review of disaster chain mechanisms

Abstract

Urban engineering systems (UESs) are highly interconnected, forming complex dependencies that render them vulnerable to cascading failures during disasters. While existing studies have explored specific aspects of disaster chains in UESs, a synthesized framework for understanding their interdependencies, data acquisition challenges, and methodological limitations remains underdeveloped. This paper addresses this gap by conducting a systematic review of UES disaster chains, beginning with the definitions of disaster chains from different academic perspectives, common types of urban disaster chains, namely earthquake, flood, fire, freezing and ground subsidence disaster chains, as well as the interdependency of UES. Furthermore, three identification methods of disaster chains are summarized, namely based on historical disaster data, expert experience, and natural language processing (NLP). Moreover, five analysis methods of disaster chains are summarized, including those based on Bayesian networks, complex networks, numerical simulation, scenario simulation and remote sensing, with comparison of their applicability, advantages, limitations and complexity. The benefits and drawbacks of each approach are clearly illustrated. The paper concludes by discussing the limitations in the current literature and suggests that future research may utilize new technologies to facilitate data analyzing process, conduct cross-regional studies, and focus on integrating socio-economic factors for disaster-related decision-making support.